psych 350 exam 2 essays

Describe the process of NHST, tell the (five) possible outcomes and tell the likely reasons for each. (Be sure to tell what this acronym means.)

NHST (Null Hypothesis Significance Testing) involves obtaining a sample, collecting data, running statistical analyses, obtaining a summary statistic and p-value, and deciding whether to retain or reject H₀. Correctly retaining or rejecting H₀ occurs with good sampling and measurement. Type I and III errors occur with poor sampling, poor measurement, confounds, or bad luck. Type II errors occur with the same issues plus too small a sample size.

Tell when to use each type of ANOVA, the possible research hypotheses for this statistical model, and when ANOVA can be used to test each type of Research Hypothesis (attributive, associative and causal).

ANOVA compares mean DV scores across IV conditions. BG ANOVA is for between-groups designs, WG ANOVA is for within-groups designs. Research hypotheses are G1>G2, G1=G2, or G1<G2. ANOVA cannot test attributive hypotheses, can test associative hypotheses when data are appropriate, and can test causal hypotheses when the design is experimental and confound-free.

Tell when to use Pearson's correlation, the possible research hypotheses for this statistical model, and when correlation can be used to test each type of Research Hypothesis (attributive, associative and causal). .

Correlation tests for a linear relationship between two quantitative variables. Research hypotheses include +r, r=0, and -r. It cannot test attributive hypotheses but can test associative hypotheses when data are appropriate. It can test causal hypotheses when data come from a true experiment without confounds

Tell when to use Pearson's Chi-square, the possible research hypotheses for this statistical model, and when Chi-square can be used to test each type of Research Hypothesis (attributive, associative and causal).

Chi-square tests for a pattern of relationship between two qualitative variables. Research hypotheses include a specific pattern or no pattern between variables. It cannot test attributive hypotheses, can test associative hypotheses with appropriate data, and can test causal hypotheses if data come from a true experiment without confounds.

Compare and contrast the 'interesting pairs' of the four bivariate data analysis models we are working with.

All bivariate tests examine if sample relationships reflect population relationships. Pearson’s correlation and Chi-square both test associations (quantitative vs qualitative). BG and WG ANOVA both test mean differences (between- vs within-groups). BG ANOVA and Chi-square are for between-groups designs (quantitative vs qualitative DVs). Correlation and WG ANOVA both involve two quantitative variables.

Respond to and describe the statement, "Rejecting the null hypothesis guarantees support for the research hypothesis."

Rejecting H₀ does not guarantee support for RH:. The RH: might be the same as H₀:, or results could contradict RH:. Retaining H₀ supports RH: only if RH: is identical to H₀.

Describe effect size estimates, tell how they are related to significance tests, and the information they provide that is not provided by significance tests. .

Effect size estimates (r) measure the strength of a relationship in correlation, Chi-square, and ANOVA. Larger effect sizes make H₀ rejection more likely. They provide a single-number summary of relationship strength and allow comparison across studies—information not given by significance tests

What is meant by "statistical power" and what is the advantage if our research has lots of it? Describe how power analyses are conducted and how they can inform our statistical decisions.

Statistical power (sensitivity) is the ability to reject H₀ when a true relationship exists. High power reduces Type II errors. A priori power analysis estimates needed sample size before data collection, post hoc power analysis estimates the probability of a Type II error after retaining H₀. Power helps plan and interpret statistical decisions.